Well if you’re downloading, copying or creating large models that are several hundred GBs you’re going to want a normal drive. SMRs have a small staging area and once that is full it has to start re-ordering the data on the platters. Once your drive is in the process of re-ordering your write speeds are going to make it look like a failing floppy disk. I had a large file copy operation (>1TB) to a RAID pool of SMRs take like 16 hours. And I also found out that my backup drive is SMR because it took several days to do a full backup from scratch, which caused me to look up its detailed specs.

It always starts out looking great but eventually the staging area will get full and then your CPU will spend most of its time twiddling its thumbs until a chunk of staging area becomes available again; Repeat until operation is done. The greatness of Shitty Magnetic Recording.

If you want to know what recording method your drive uses, you grab the model number from:

# smartctl -a /dev/YOURDRIVE | grep "Device Model"

and look that up in a search engine. It should either lead you to a Data sheet or the manufacturers website where they list the specifications.

If durable and large SSDs were more affordable where I am I’d slowly replace all my spinning rust. But right now HDDs are overall still the better option for me, at least for mass storage.

You must be new to FLOSS, because this isn’t a real secret. In fact, he’s pretty much known for being an asshole. He can still be pretty blunt these days, but he was a lot worse in the past. Also why does it matter? I genuinely don’t understand the point of this post.

(and an HDD that size costs almost nothing now)

Sure, but you really have to watch out what kind of hard drives you’re buying. There are a lot of SMR drives out there that are sold as regular drives and the only way to tell is to look through their data sheets. I find that “regular” HDDs (CMR/PMR) cost more now than SMR drives of similar capacity and spindle speed (probably because nobody wants them lol).

DW. in like two years from now, companies will start throwing out similar machines. Just keep an eye on second-hand markets and dumpsters.

Anyone who’s been following Kernel development for a while knows that big tech has been all over Linux for a while, and I’d count the relatively recent inclusion of rust in the Kernel as a major shift in policy. Linux has sponsors, not donators. Also since the Linux Foundation and most of their infrastructure is based in burgerland, they have to follow US law.

I’m not trying to be a doomer, but Linux is far less autonomous than people like to think. However, since it’s FOSS anyone can just modify it and rip out the junk. And it’s still the only FOSS Unix-like that supports most modern Hardware, which probably makes up most of its value.

There is a project I’ve discovered recently which is similar to GPT4All, except you can throw multiple GPUs at the workloads (and yes it can use Vulkan): https://github.com/LostRuins/koboldcpp

I haven’t messed much with it but it builds and works fine on Linux. The only thing I don’t like is that the source tree has a bunch of Windows binaries in it.

But most importantly, it won’t work in the end. These scraping tech companies have much deeper pockets and can use specialized hardware that is much more efficient at solving these challenges than a normal web browser.

A lot of people don’t seem to be able to comprehend this. Even the most basic Server Hardware that these companies have access to is many times more powerful than the best Gaming PC you can get right now. And if things get too slow they can always just spin up more nodes, which is trivial to them. If anything, they could use this as an excuse to justify higher production costs, which would make resulting datasets and models more valuable.

If this PoW crap becomes widespread it will only make the Internet more shitty and less usable for the average person in the long term. I despise the idea of running completely arbitrary computations just so some Web Admin somewhere can be relieved to know that the CPU spikes they see coming from their shitty NodeJS/Python Framework that generates all the HTML+CSS on-the-fly, does a couple of roundtrips and adds tens of lines of log on every single request, are maybe, hopefully caused by a real human and not a sophisticated web crawler.

My theory is people like to glaze Anubis because it’s linked to the general “Anti-AI” sentiment now (thanks to tech journalism), and also probably because its mascot character is an anime girl and the Developer/CEO of Techaro is a streamer/vtuber.

I want an AI, for example, to analyze the material conditions of a country and plan a Communist revolution for me. Can I have that? Will capitalism produce this for me?

Unironically, I think what’s more likely to happen is that once the compute resources become affordable to the average working class people then someone will inevitably start a FOSS project like that. I have the feeling even if some private company manages to make an AI that could be used for things like planning revolution they will probably train it in ways to make it incapable of doing these things.